The care "of critically ill patients is a common, expensive, and growing part of American medicine. Some hospitals provide higher quality care to these patients than do other hospitals. Every year approximately 10,000 patients die from acute myocardial infarction who might have been saved had they been at a higher performing hospital. Similarly large numbers of patients die potentially avoidable deaths after mechanical ventilation. Given these large variations between hospitals in the quality of their care, there is growing interest in developing a formal system to transfer patients to higher quality hospitals. While there are calls for a formal system for transferring patients, an unplanned system has already evolved for the transfer of critically ill patients. We know critical care transfers are common, but we know very little abput the network over which they are transferred. In particular, we do not know whether the network directs patients towards higher quality hospitals. The global hypothesis of this proposal is that the critical care transfer network is a coherent structure, not just an aggregation of random processes. This structure, we hypothesize, is directly related to hospital organizational characteristics, impacts patient outcomes, and is amenable to intervention. We will perform three Specific Aims. First, we will map and characterize the network structure. Second, we will examine the association between a hospital's centrality in the network - that is, the degree to which the network moves patients towards to the hospital - and hospital organizational characteristics and capabilities. Third, we will determine the relationship between a hospital's network centrality and outcomes of its patients: We will use as our model system 30-day mortality after one of the most common disease processes requiring critical care: acute myocardial infarction. Data for this project will be derived primarily from the Medicare claims. This research is directly relevant to the goal of providing high quality care to patients with acute illness, particularly acute cardiac and pulmonary derangements. This proposal investigates whether the existing network of critical care transfers can be used as a foundation to improve the care of patients by moving patients to hospitals that provide better care. (End of Abstract)